spateo.external.CAST.CAST_Mark

Functions

train_seq(graphs, args, dump_epoch_list, out_prefix, model)

The CAST MARK training function

delaunay_dgl(sample_name, df, output_path[, if_plot, ...])

Module Contents

spateo.external.CAST.CAST_Mark.train_seq(graphs, args, dump_epoch_list, out_prefix, model)[source]

The CAST MARK training function

Parameters:
graphs List[Tuple(str, dgl.Graph, torch.Tensor)]

List of 3-member tuples, each tuple represents one tissue sample, containing sample name, a DGL graph object, and a feature matrix in the torch.Tensor format

args model_GCNII.Args

the Args object contains training parameters

dump_epoch_list List

A list of epoch id you hope training snapshots to be dumped, for debug use, empty by default

out_prefix str

file name prefix for the snapshot files

model model_GCNII.CCA_SSG

the GNN model

Returns:

returns a 3-member tuple, a dictionary containing the graph embeddings for each sample, a list of every loss value, and the trained model object

Return type:

Tuple(Dict, List, CCA_SSG)

spateo.external.CAST.CAST_Mark.delaunay_dgl(sample_name, df, output_path, if_plot=True, strategy_t='convex')[source]